Exposé à Toulouse 1, sur l'estimation nonparamétrique de quantiles.

In this talk we propose several nonparametric estimators of quantiles based on Beta kernel and applied to transformed data by the generalized Champernowne distribution initially fitted to the data. A Monte-Carlo based study will show that those estimators improve the efficiency, not only for light tailed distributions, but mainly for heavy tailed, when the probability level is close to 1.Another application will be seen, on portfolio optimization in the mean-VaR context.